Consideração dos efeitos de potência reativa na análise de perdas elétricas em sistemas de potência

Abstract

This work discusses the problem of active power loss reduction in power systems by means of reactive power redispatch. The circulation of reactive power in the network causes active power loss in the system. Then, a reduction in active power loss can be achieved with a proper adjustment of reactive sources in the system. When it comes to generators reactive power redispatch, the physical location of these generators in the system must be taken into account, because of the local nature of the reactive power. So, depending on system configuration, the minimization of the active power loss cannot be achieved by means of the generators redispatch. This can also occur through incorrect choice of generators to participate in the redispatching. Therefore, this work proposes a simple method based on sensitivity analysis to determine the generator most likely to participate in the reactive power redispatch, aiming at reducing active power loss in the system or in an area of interest. The use of the tangent vector is also presented for this purpose. The problem of generators reactive power redispatch is solved by incorporating a constraint in a modified Jacobian power flow, whose solution is obtained by using the Newton-Raphson method. The considered constraint is the active power loss in the area of interest. . In this work, the area of interest is the critical area, since the system security is closely related to active power loss at system’s critical area. Since the loss reduction problem in power systems is usually approached through optimization techniques, a comparison between the solution of redispatching method and the solution obtained by means an Optimal Power Flow program is also performed. One advantage of the methodology presented in relation to the Optimal Power Flow is the ease of implementation in any program of power flow. A computer program for power flow that implements the methodology for identifying and redispatching generators was developed in Matlab®. The results are presented through two systems: an academic system and a real Brazilian system.